4 Transmission issues: diffusion and development of technology
4.3 Consequences for technology development
CEUeTDCollection
125
CEUeTDCollection
126 governments. Also, we have seen in Section 4.3, the resulting developmental alliance between ECE states and the multinationals, which is responsible for the growth of local production, is supported by the multinationals’ private coordination networks, as well as by the semi-public organisations of foreign chambers of commerce. The existence of this transnationalized institutional framework only reinforces the exclusion of domestic companies, but it also adds a degree of stability to the localized clusters of foreign firms, reinforcing network and agglomeration effects.
In this context, it is hard to put forth a convincing defence of domestic capital, especially where, like in East Central Europe, it also commands very little political leverage or popular support. Compared to the multinationals, domestic companies usually also offer much lower wages, and are not necessarily a more stable source of employment. So far, there is little evidence for the common argument that foreign firms are more likely to relocate than the nationally owned ones: relocations in automotive industry have been few and far apart, and there is no reason to expect domestic companies to react differently to the cost pressures. In fact, Videoton in Hungary was among the first firms to move the most labour-intensive jobs out of the country in the face of radical restructuring in the industry (Radosevic & Yoruk 2001).
While it is probably unlikely that they would take their entire operations out of the home country – if for no other reason, than for the lack of funds – as argued above, their greater
“commitment” to local production had so far been easily counterbalanced by the higher probability that they would downsize or disappear altogether. However, there is one aspect in which the lack of large domestic firms might eventually have detrimental consequences for the future development of regional manufacturing: the ability to move beyond improvements in production to development of new products and processes – in other words, from technology transfer to technology development.
In the model outlined in Chapter 2, I defined technology transfer as appropriation of modern technologies by a developing country, in order to increase local industry’s productivity and make its output internationally competitive. This transfer has been considered a major
CEUeTDCollection
127 challenge by most theorists of late development, who understood it to mean transfer of technology between companies, i.e. from foreign to local firms. We have seen, however, that in East Central European hyper-integrationist variety of late development this challenge all but disappeared, because the bulk of transfer took place within foreign firms, through establishment of local affiliates. Technology transfer to domestic firms was much more problematic, but due to the large-scale influx of foreign capital which substituted the missing production capabilities, its economic impact appears to be marginal.
On the other hand, there is a concern that this rapid transplantation of foreign production networks, however efficient, might have detrimental consequences on the economy’s ability to move towards more knowledge-intensive activities, such as research and innovation. Indeed, marginalization of domestic companies as described above had a particularly negative effect in this area. Even for the most successful firms, the priority for a long while was to acquire up-to-date foreign technology, not to develop their own, and surviving in the lower ranks of the multinationals’ value chains often meant abandoning complex research activities (Radosevic & Yoruk 2001; Pavlínek 2012). Where the local firms were taken over by their foreign competitors, they usually lost existing R&D capacities, which were transferred to the MNC’s home country (Pavlínek 2012). Flexibility, which remains among the main advantages of the local producers, can also be detrimental to their innovative capabilities: as we have seen in the case of Matador, a timely switch into a new product niche to avoid the price squeeze in an over-concentrated market might mean abandoning decades-old accumulation of research experience.
But if the domestic firms are unlikely to serve as the engine of technology development, could their weakness be sufficiently compensated by foreign investment, like we have seen in the case of production? In other words, is the integrationist path inevitably a no-technology path, as Amsden had feared, or can we expect the process of fragmentation of global value chains to also bring the knowledge-intensive activities to the peripheries?
CEUeTDCollection
128 To date, the recent literature has no conclusive answer to this question. According to the more optimistic arguments, new information technologies have facilitated rapid internationalization of corporate R&D, and as research and design comes to account for an ever larger share of the total product cots, they will become a likely target for outsourcing (Narula &
Guimon 2010; Bruche 2009; Karlsson 2006; Moncada-Paternò-Castello et al. 2011; Criscuolo 2009). Others are more sceptical. A survey by UNCTAD (2005b), for instance, has shown that internationalization is limited to certain industries and countries, and that although more R&D investment has been going abroad, most of it is concentrated in other developed countries, or in the largest developing countries where market conditions require extensive product adjustments (also Puga & Trefler 2010).
The evidence available for automotive industry in Europe generally confirms that internationalization pressures described in Section 4.1 only reinforced centralization of higher-level competencies. Research and development in automotive industry is fairly capital intensive and usually concentrated in the largest firms. In spite of extensive decentralization which took place in the recent years, final vehicle manufacturers are still responsible for around half of all corporate R&D in the industry (ILO 2005), which is mainly performed in their home countries, or in other developed countries (Calabrese 2001, Lung 2004). Since the automotive industry is still regionally organised, some R&D goes to the countries at the centre of other regions of the Triad, to ensure that the locally produced vehicles are adapted to the consumer demands and public regulation (e.g. in the case of American investments in Europe, whose research and development activities are mainly located in Germany). Exceptionally, these transfers of capabilities can also benefit less central locations, especially when they come from the newer, less established and more price-conscious firms: thus, for instance Visteon, a major Tier-1 supplier which was spun off from the Ford’s component division in the late 1990s decided to establish its European lighting centre in the Czech Republic (Pavlínek 2012).
The process of devolution of research and development to the suppliers has also been quite limited: the top 100 firms now account for three quarters of all R&D investment in the
CEUeTDCollection
129 sub-sector (Dannenberg and Burgard, c.f. Pavlínek 2012). Also, in spite of the spread of information technologies and “virtualization” of exchanges, the collocation pressure in research and development are very strong, and most vehicle manufacturers require their key suppliers to situate research capacities close to those of the lead firm. In consequence, central regions’ share of R&D investment and employment even increased in the course of late 1990s and 2000s (Lung 2004; Pavlínek 2012; also Criscuolo & Narula 2005). An illustrative example comes from one of the most successful Spanish suppliers, the Ficosa Group, which in the early 1990s faced the pressure of rapidly internationalizing and expanding its capacities or being pushed down in the value chain. A traditional supplier of Ford in Spain, Ficosa had to keep up with its customer’s plans to integrate all European production, and it consequently invested in research and development centres abroad, in the UK and Germany, in order to be able to profit from proximity in the decisions on design and development of new products. (Durán Herrera 1996;
Margalef Llebaría 2005).
In East Central Europe, investment in research and development has certainly not kept pace with the development of production, and the share of R&D in total expenditure and employment remains below the levels of the “core” producers (Figure 4.2, see also Pavlínek et al. 2009). The measure of R&D output, such as the number of registered patents, shows a very similar picture. In the course of the 1990s, the number of automotive patents skyrocketed in France and Germany, but remained stagnant in most other European countries, confirming the above mentioned tendency towards concentration. In the ECE, all countries registered a small increase over the course of 2000s, but remained at very low levels compared to the European core. The number of annual patents per 1000 workers in the late 2000s was less than 1 in all countries of the region, although the Czech Republic came closest with 0.8. By comparison, France and Germany record on average 10 new patents per thousand workers, but this pattern seems to be strongly correlated with the location of corporate headquarters. Other West European countries, which are far more developed and have more experience with
CEUeTDCollection
130 international automotive production – such as Belgium or Spain – show similarly low levels of patent activity in this industry (Figure 4.2).
This is not to say that some R&D has not found its way into the regional automotive cluster, but its real impact is very difficult to judge. Pavlínek et al., for instance, document around 40 research and development centres in the region, most of them established by the international Tier-1 suppliers. They also show that such investments have increased in the recent years: more than half of these centres were established after 2004 (Pavlínek et al. 2009).
Most of them are adjacent to the local manufacturing plants, and focus on development and adaptation of components for the models produced specifically in the region (Domanski et al.
2008). Although small in international terms, given the demise of large domestic firms these investments now constitute most of automotive R&D performed in the region: in the most comprehensive survey to date, Pavlínek (2012) finds that almost 80% of all R&D employment in
Figure 4.2R&D in automotive industry in ECE and selected WE countries17
17 The patents were compiled based on priority numbers, which correspond to the application number of
the document in respect of which the priority is claimed. The figure includes patents registered under IPC codes B60B, B60D, B60G, B60H, B60J, B60K, B60L, B60N, B60P, B60Q, B60R, B60S, B60T, B62D, E01H, F01L, F01M, F01N, F01P, F02B, F02D, F02F, F02G, F02M, F02N, F02P, F16J, G01P, G05D, G05G, based on Eurostat NACE-IPC concordance tables. Since the IPC codes describe technological content of innovation and are not directly comparable to the classifications based on economic activity, the patents above cover a more broadly defined vehicle industry (e.g. including agricultural, construction and military vehicles).
0.0 2.0 4.0 6.0 8.0 10.0 12.0
0 5 10 15 20 25
CZ HU PL SK DE FR IT UK BE ES PT
R&D as % of value-added R&D as % of employment Patents per 1000 workers
Note: averages for 2005-2007
Source: own calculations based on Eurostat, EPO Worldwide database
CEUeTDCollection
131 the Czech Republic is in the foreign-owned firms, even if the largest investor, Škoda, is excluded from the count. At the same time, these figures indicate the scale of dependence on a few large investors: Škoda alone accounts for more than 75% of R&D expenditures in the Czech industry (see also Narula & Guimon 2010). The content of these activities is even more difficult to judge then the scale. Much of it consists of technical support to production, and the differentiation is further compounded by high levels of automation in vehicle manufacturing, inducing a shift from manufacturing to programming and maintenance. The European Union regulation on state aid, which prohibits assistance to capital investments but allows generous support to research and development has likely led to some inflation of the concept of “R&D”, pinning this prestigious title on even the relatively simple technical and business support activities (see also Pavlínek 2012).
Some of this discrepancy becomes visible if we look more closely into the source of automotive industry patents. Clearly, this is an equally imperfect measure, as the number of patents does not necessarily say much about their market value or the extent of innovation.
Nevertheless, if we exclude private individuals and restrict the analysis to those entities which in the period 1990-2010 registered at least ten patents, as an indicator of sustained innovation activity, the picture is a lot less encouraging. Over the two decades under consideration, there were altogether 55 such entities in Central and Eastern Europe, three in Slovakia and Hungary, 18 in the Czech Republic and 32 in Poland. Of these, only 9 were foreign owned, and 26 were universities or public research institutes. In Hungary, all three such entities were domestically owned vehicle and component manufacturers, and two of them (Csepel Auto and Ikarus) went bankrupt in the course of the 1990s, allowing us to trace the decline of innovation activity in Hungary directly to the disappearance of even these second-class national champions. Similarly, in Slovakia the highest number of automotive-related patents was registered by Matador, all of them in the rubber industry which the company in the meantime abandoned.
Among the foreign-owned companies with a consistent patent activity, six were to be found in the Czech Republic, and three in Poland. Only one of them – Škoda Auto – is a vehicle
CEUeTDCollection
132 manufacturer, which shows that even with the strong pressure for R&D collocation, investments by supplier might be easier to attract than those by the carmaker. However, without a carmaker leading the way, they are unlikely to be very extensive. A local lead firm that acts as a hub creates demand for R&D from localizing suppliers, and allows local suppliers to prove their ability to supply such R&D before they can extend their services to other manufacturers. In the Czech Republic, Škoda has been able to play this role because its separate brand within the Volkswagen concern allows it autonomous design and development functions for at least some parts of the vehicle (Dörr & Kessel 2002; also Pavlínek 2012). Consequently Czech Republic has been the most successful country in attracting FDI in research and development, and the patent analysis reveals extensive innovation activity not only by Škoda, but also by Visteon, Robert Bosch and Siemens, among others. At the same time, Czech Republic also has the most patent activity among domestic suppliers. This, and the experience of some more successful countries of the relative “periphery”, such as Spain, suggests that far from being substitutes, research and development activities by foreign and domestic companies are in fact the result of the same set of favourable conditions, not the least of which is the existence of a relatively autonomous local lead firm.
It would thus appear that successful local development of technology in late starters requires a degree of insulation, or other kinds of regulatory pressure that favour innovation over import of technology. There is, however, a degree of trade-off between this focus on development of local capabilities and rapid increase in competitiveness driven by substitution.
Exceptionally, demand for local investments in knowledge-intensive activities can arise even in absence of explicit regulatory intervention, for instance as a consequence of a large and specific market (as is the case in some BRICs), or from a historical accident and path dependency, such as has produced Škoda’s unique position within the Volkswagen concern.
If the costs of insulation are too high, and technology development is unlikely to arise in response to demands of the local markets (either by final consumers or by local assemblers), an alternative for the host state would be to supply the right institutional incentives to convince
CEUeTDCollection
133 foreign firms to undertake some research and development in the location. This is a pattern familiar from R&D investments between developed countries, as when the foreign firms set up research centres in the Silicon Valley in order to profit from the local talent and networks. In the case of developing countries, however, this strategy is vulnerable to the same kind of circular reasoning we encountered with respect to spillovers: in order for the foreign capital to engage in technology development, the host state must create an institutional infrastructure conducive to production of cutting-edge technologies.
In its absence, the most common strategy among the East Central European governments is to persuade the multinationals to undertake at least some research, hoping that these seeds will eventually expand into full-scale innovation clusters. Low wages can help to attract the more labour-intensive research activities, such as testing, and these are indeed on the rise in a number of industries in ECE (Domański et al. 2008; Pavlínek 2012), but they are often performed in isolated outposts and are considered quite footloose (Gunther et al. 2009).
In addition to that, national investment agencies are aggressively advertising both real and symbolic engineering potential of the region (references to the late 19th and early 20th century innovators are a staple fare of all promotional materials), and the governments are offering generous subsidies to all innovation-related activities. Nevertheless, these essentially come in addition to and not instead of, traditional support to capital investments. Even with EU pressure
Figure 4.3 State aid in ECE and EU-15 per type of activity, 2004-2011 (%GDP)
0 0.1 0.2 0.3 0.4 0.5 0.6
CZ HU PL SK CEE EU15
Training Environment Entrepreneurship R&D
Regional
Note: "Regional" includes aid to capital investments and employment, "Entrepreneurship" aid to SMEs and provision of risk capital for new undertakings.
Source: Own calculations based on EC State Aid Scoreboard 2012.
CEUeTDCollection
134 to steer state aid from mere employment to higher value-added activities, between 2004 and 2011 the bulk of public subsidies to industry in ECEs (between 50% in CZ and 75% in Hungary) still went to support regional development and employment, compared to about 30% in EU-15 (Figure 4.3). Although ECEs give on average more subsidies than the old EU members relative to GDP, they are less likely to spend them on research-related projects, and the level of business expenditure on R&D is generally among the lowest in the OECD world (OECD 2012).
While there are not many ways for the ECE governments to force foreign firms to discover the advantages of local skill supply, they have also been trying to promote R&D from the other side – by either helping or bullying the national research institutions into partnerships with the multinationals. Starting in 2000, Hungary called on the universities to establish so-called “Cooperative Research Centres”, and provided special funding to those institutions which managed to assemble a consortium of private firms for joint research. The scheme was a relative failure, however, and only yielded a number of short-term projects, but in 2004 the Hungarian government launched its successor, this time with more generous funding and with a clause that any partnership established under the programme should last for at least ten years (Inzelt 2008). Nineteen regional Knowledge Centres were subsequently established, three of them in the automotive industry. The most successful one, Advance Vehicle Control Knowledge Centre, is a partnership between the Budapest Institute of Technology (BME) and Knorr-Bremse and Thysen-Krup, both of which already have large research centres in Hungary. A much less gentle attempt to create more linkages between public and private research was demonstrated by the Czech government, which in 2009 announced that its funding to the academy of sciences will be cut by 45% over the next three years. In exchange, research institutions were urged to focus on more applied and commercial research and find alternative financing from the private sector (Government Office of the Czech Republic 2009).
So far, even this approach has yielded modest results. Most leading automotive firms have some cooperation with local universities: Volkswagen in Slovakia cooperates with technical universities in Bratislava and Kosice, and Audi in Hungary has a long standing
CEUeTDCollection
135 cooperation with the local technical college, where it has helped to establish a special department on internal combustion engines. However, apart from the already mentioned knowledge centre at BME and some joint projects between universities of Prague and Liberec and a number of large Tier-1 suppliers (Honeywell, Bosch), these are not strictly research partnerships – rather, they involve occasional lectures, equipment donation and science fairs, and serve rather as information and recruitment channels than as sites of technology development.
CEUeTDCollection
136